NumeroLogic: Number Encoding for Enhanced LLMs' Numerical Reasoning
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866916411323973632 |
|---|---|
| author | Schwartz, Eli Choshen, Leshem Shtok, Joseph Doveh, Sivan Karlinsky, Leonid Arbelle, Assaf |
| author_facet | Schwartz, Eli Choshen, Leshem Shtok, Joseph Doveh, Sivan Karlinsky, Leonid Arbelle, Assaf |
| contents | Language models struggle with handling numerical data and performing arithmetic operations. We hypothesize that this limitation can be partially attributed to non-intuitive textual numbers representation. When a digit is read or generated by a causal language model it does not know its place value (e.g. thousands vs. hundreds) until the entire number is processed. To address this issue, we propose a simple adjustment to how numbers are represented by including the count of digits before each number. For instance, instead of "42", we suggest using "{2:42}" as the new format. This approach, which we term NumeroLogic, offers an added advantage in number generation by serving as a Chain of Thought (CoT). By requiring the model to consider the number of digits first, it enhances the reasoning process before generating the actual number. We use arithmetic tasks to demonstrate the effectiveness of the NumeroLogic formatting. We further demonstrate NumeroLogic applicability to general natural language modeling, improving language understanding performance in the MMLU benchmark. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_00459 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | NumeroLogic: Number Encoding for Enhanced LLMs' Numerical Reasoning Schwartz, Eli Choshen, Leshem Shtok, Joseph Doveh, Sivan Karlinsky, Leonid Arbelle, Assaf Computation and Language Language models struggle with handling numerical data and performing arithmetic operations. We hypothesize that this limitation can be partially attributed to non-intuitive textual numbers representation. When a digit is read or generated by a causal language model it does not know its place value (e.g. thousands vs. hundreds) until the entire number is processed. To address this issue, we propose a simple adjustment to how numbers are represented by including the count of digits before each number. For instance, instead of "42", we suggest using "{2:42}" as the new format. This approach, which we term NumeroLogic, offers an added advantage in number generation by serving as a Chain of Thought (CoT). By requiring the model to consider the number of digits first, it enhances the reasoning process before generating the actual number. We use arithmetic tasks to demonstrate the effectiveness of the NumeroLogic formatting. We further demonstrate NumeroLogic applicability to general natural language modeling, improving language understanding performance in the MMLU benchmark. |
| title | NumeroLogic: Number Encoding for Enhanced LLMs' Numerical Reasoning |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2404.00459 |